Diet quality and nutrient distribution while using glucagon-like-peptide-1 receptor agonist: A secondary cross-sectional analysis

Abstract

Background

With the rise of glucagon-like-peptide-1 receptor agonist medications (GLP-1RA) for obesity treatment, understanding diet quality can be a valuable tool for providing evidence-based nutrition guidance. However, there is limited data on dietary intake during GLP-1RA treatment. Thus, we analyzed diet quality and nutrient timing while using GLP-1RA.

Methods

This was a secondary analysis from a previous cross-sectional online survey questionnaire study involving adults currently using GLP-1RA for weight reduction (N = 69, 49.6 ± 12.3 years old, 35.9 ± 9.1 kg/m2). Three-day food records were analyzed using the Healthy Eating Index (HEI), a validated score that indicates overall diet quality. The 13 HEI components were scored from average 3-day food records and calculated using 95 % confidence intervals (CI). A Bonferroni correction applied significance accepted at p = 0.0038. Additionally, 95 % CI were calculated for calories, macronutrients, and fiber intake reported for breakfast, lunch, dinner, and snacks.

Results

A 95 % CI revealed a total HEI score of 54 ± 12 (51.4, 57.3), significantly below the HEI goal (p < 0.0038). All components, except added sugars, were significantly under the max score. There was no significant difference for HEI scores based on duration of GLP-1RA use. The largest number of calories were consumed at dinner, averaging 649 compared to 538, 392, and 391 calories at lunch, breakfast, and snacks, respectively. Further, 40 % of the total daily protein intake occurred at dinnertime.

Conclusion

Within the sample of patients using GLP-1RAs, dietary quality was suboptimal for fruits, vegetables, whole grains, seafood and plant proteins, dairy and fatty acids. Future research is needed to determine if HEI scores change before, during, and after GLP-1RA treatments and nutrient timing.

Graphical abstract

Highlights

• •Diet quality was suboptimal with GLP-1RA use with no differences in Healthy Eating Index scores based on treatment duration.
• •Calorie and protein intake were disproportionately consumed at dinner compared to other meals.
• •Results highlight a need for future studies aimed at improving diet quality and nutrient timing with GLP-1RA use.
• •Providers and dietitians play a key role in guiding patients to improve diet quality and nutrient timing with GLP-1RA use.

Abbreviations:

GLP-1RA

Glucagon-like-peptide-1 receptor agonists

HEI

Healthy Eating Index

1. Introduction

Diet quality has been a focus of nutrition research for decades to better understand how to promote health and prevent disease. High-quality dietary intake is inversely correlated with several chronic diseases as well as body weight [1]. Focusing on diet quality as an intervention should be a primary focus to promote better health with the prevalence of obesity among U.S. adults at 41.9 % [2]. Recent advancements in weight reduction medications, such as glucagon-like-peptide-1 receptor agonists (GLP-1RA), have dramatically changed the landscape of obesity treatment. These medications mimic the naturally occurring GLP-1 hormone at much higher potencies, resulting in slower transit time of gastric emptying, appetite control, blood glucose regulation, and weight reduction. A once-weekly GLP-1RA injection provides approximately 14.9 % weight reduction, with some as high as 20 % compared to an average of 5–7 % using traditional diet and lifestyle modifications with concerns related to fat-free mass [3,4]. Proper nutrition, such as nutrient-dense dietary patterns, and exercise interventions are needed along with GLP-1RA treatments to promote long-term health and minimize loss of muscle mass [5,6].

Comprehensive obesity treatment with GLP-1RA should focus on adequate nutritional and dietary quality to support healthy weight reduction outcomes. Very few randomized clinical trials have reported dietary intake while using a GLP-1RA. The currently available evidence on dietary intake is limited to average energy consumption and macronutrient intake, mostly from laboratory environments [7]. In a 14-week clinical trial using 24-h ​food recalls, participants using 1.8 mg/d liraglutide injection reduced daily energy intake on average by 294 calories over the study period [8]. Another study evaluated ad libitum intake during week 12 of Semaglutide treatment and found a 24 % reduction in total calories (approximately 725 calories) compared to placebo [9]. Gibbons et al. found a 38.9 % lower ad libitum total energy intake (around 1200 calories) compared to placebo during week 12 of oral semaglutide treatment. However, none of these studies reported specific diet quality and composition [10]. Publications across the landscape of obesity interventions can provide some evidence for the nutritional status and insights for the GLP-1RA population. For example, a previous study measuring baseline nutritional status of individuals with obesity found prevalence of micronutrient deficiencies from serum concentration for vitamin D, vitamin C, selenium, and iron [11]. Another study found 48.7 % of individuals showed at least one prevalent deficiency in key serum nutrients prior to bariatric surgery but vitamin and mineral intake from food was not recorded [12]. Prior research in individuals with obesity has shown 5–12 % lower intakes of micronutrients and higher prevalence of nutrient inadequacy compared to normal weight subjects [13]. These results suggest lower quality dietary intake as high-quality diets are often rich in nutrients.

The United States Department of Agriculture provides Dietary Guidelines for Americans (DGA), updated every five years, to inform nutrition policy and provide guidance on healthy diets [14]. These guidelines provide practical recommendations for dietary intake based on scientific evidence. However, for individuals using a GLP-1RA and reduced energy intakes, it can be challenging to meet dietary guidelines developed for a standard 2000 calorie diet. The Healthy Eating Index (HEI) is a validated measure of overall diet quality, independent of quantity, likely making it a more suitable measure of diet quality for the GLP-1RA population who have reduced calorie intake. Further, the HEI is used in NHANES (National Health and Nutrition Examination Survey) data reporting on dietary quality of the U.S. population [15], allowing a comparison to determine if individuals using a GLP-1RA have similar or different choices in overall dietary quality. Emerging evidence suggests individuals using a GLP-1RA are choosing healthier foods. Recent survey data on grocery purchases show shifts toward healthier foods with consumers using a GLP-1RA. These individuals are spending less money on high-calorie, high-sugar, and high-fat items, with a modest increase in fruit and yogurt [16]. However, it's unknown whether individuals using a GLP-1RA adopt overall high-quality dietary patterns. A recent joint advisory from four U.S. health agencies commission encourages dietary guidance on nutrient adequacy from fruits, vegetables, whole grains, lean proteins, and legumes while limiting refined carbohydrates, sugary beverages, and processed foods while using a GLP-1RA [17]. Further, the timing of nutrient intake, specifically in relation to protein distribution throughout the day, is unknown. Prior clinical evidence shows protein consumption evenly distributed throughout the day supports muscle protein synthesis better than a larger protein intake at a specific meal [18]. To our knowledge, previous research has not evaluated nutritional intake using the HEI to determine dietary quality in a GLP-1RA population for longer than a 24-h ​recall. Therefore, this secondary analysis examined data from a previous study to assess dietary quality and nutrient intake timing while using a GLP-1RA.

2. Methods

This was a secondary analysis from a previous cross-sectional online survey questionnaire study involving adults currently using GLP-1RA for weight reduction (N = 69, 49.6 ± 12.3 years old, 35.9 ± 9.1 kg/m2). This study was a secondary analysis of a cross-sectional study conducted in a group of individuals currently using a GLP-1RA living in the United States [U.S.] [19]. The study utilized various methods, including a 3-day food record to measure nutritional intake. The methods are briefly described in this paper, while detailed information about the methodology and enrollment can be found in the previous publication [19]. The study was conducted September 2024 to October 2024, and all participants provided written consent upon enrollment. The research procedures were approved by an Institutional Review Board and all participants provided informed consent prior to voluntary participation [BRANY IRB-Manager, New York, USA; Approval date 16 August 2024].

2.1. Three-day food record

A 3-day food record was collected via the Automated Self-Administered 24-Hour Dietary Assessment Tool (ASA24) [20]. The Automated Self-Administered Dietary Assessment Tool, version 2024 [ASA24] software developed by The National Cancer Institute in the United States for collecting detailed information about an individual's dietary intake [21,22]. The ASA24 software was created to improve validity by collecting 24-recalls or food records electronically. Dietary assessments used in research have inherent strengths and limitations [23], however, the 3-day food record was selected to help identify key knowledge gaps related to dietary intake among current patients using GLP-1RA [7]. Participants were instructed to record their food intake for three consecutive days, starting on the first day of their weekly injection, to standardize dietary assessment across the GLP-1RA sample. Given the current lack of evidence on how GLP-1RA administration influences day-to-day food intake, this approach was selected to ensure consistency and minimize variability in data collection. To our knowledge, it's unknown how GLP-1RA effects dietary intake fluctuations throughout the week and the duration of treatment. Participants were provided with a protocol on how to record their food logs to improve accuracy. The protocol included training on system navigation, food selection process, portion size estimations, reviewing and editing food entries, and how to record more complex meals.

2.2. Primary outcomes

In this secondary analysis, we aimed to 1) Analyze dietary quality using the HEI, 2) Compare HEI scores across individuals using GLP-1RA for different durations, and 3) Explore patterns of nutrient timing and meal composition.

The HEI is a validated score that indicates overall diet quality and has demonstrated construct validity, reliability and criterion validity and is consistent with other dietary indices [24,25]. The HEI scoring system uses a density approach and separates dietary quality from quantity for 13 components. Nine components are labeled as adequacy, which represent food groups that are encouraged, where the higher score reflects higher intakes. The remaining four components are moderation, which represent food groups recommended to limit. For the moderation components, higher scores reflect lower intakes. Twelve components of the diet are calculated as a food group per 1000 calories in the total mix of foods. Whereas the fatty acids component is scored as a ratio of unsaturated and saturated fatty acids. The HEI 2020 was updated to reflect the DGA 2020–2025; however, the calculations remained the same. To calculate the HEI, components of the diet are scored from 0 to 5 or 10, where 5 and 10 are the maximal points. Intakes between the minimum and maximum standards are scored proportionately. The sum of all 13 components is the final HEI score. The higher the HEI score, the closer food choices align with the DGA with a maximal score of 100. A population ratio method was used to calculate the total intake of dietary constituents followed by the appropriate ratios for each HEI component [26].

2.3. Statistical analysis

Data analyses were performed using the Statistical Package for Social Sciences 26. Descriptive statistics [means, standard deviations] for all participant characteristics and the HEI were calculated. The primary analysis was 95 ​% confidence intervals (CI) for the 13 components of the HEI compared to the respective maximum HEI score, with significance accepted at p ​= ​0.05. For the total HEI score, a Bonferroni correction was applied to account for multiple comparisons, adjusting the significance threshold to p ​= ​0.0038 (0.05/13 components) to control Type I errors and reduce the risk of false positives. The 95 ​% CI was calculated for each component were scored from the average 3-day food records. Further, ANOVA was used to compare HEI scores among duration of GLP-1RA use. The secondary analysis calculated 95 ​% CIs for calories, macronutrients [carbohydrates, protein, and fat], and fiber intake during breakfast, lunch, dinner, and snacks reported from the 3-day food records. Relationships between calorie and macronutrient intake and the dependent variables were explored using Pearson's correlation and analysis of variance. A post-hoc power analysis was conducted using G∗Power for a one-sample t-test with a moderate effect size (d ​= ​0.5) and a significance level of p ​= ​0.05. The analysis yielded a power of 0.99 indicating the sample size of N ​= ​69 was sufficient to detect a meaningful difference. A post-hoc G∗Power was calculated for Pearson's correlation with a moderate effect size (r ​= ​0.3) and a significance level of p ​= ​0.05. The analysis yielded a power >0.8 indicating the sample size was large enough to detect significance for potential correlations.

3. Results

3.1. Participants

Detailed descriptions of participants can be found in the original publication [19]. The data from 69 participants (14 males and 55 females) were included in the secondary analysis. Fig. 1 displays the flow of study recruitment. Key participant descriptive and frequency statistics are reflected in Table 1. All participants were currently taking GLP-1RA with the majority taking Semaglutide [n = 37, 53.6 %] and Tirzepatide (n = 23, 33.3 %). The duration of medication use among participants was as follows: 7.2 % had been taking the medication for less than three months, 29 % for four to six months, 24.6 % for seven to twelve months, and 39.1 % for over one year. Most of the participants identified as White/Caucasian (n = 57, 82.6 %) which is consistent with other findings reporting lower rates of GLP-1RA usage in Asian, African American, and Hispanic individuals by comparison [27].

Fig. 1.

Flow diagram of participants' recruitment from cross-sectional study with patients currently using a Glucagon-like-peptide-1 receptor agonist for weight loss.

Table 1.

Descriptive and frequency statistics for participant characteristics for a sample of individuals using a glucagon-like-peptide-1 receptor agonist drug.

Characteristics	Males N = 14	Females N = 55	Total Mean ± SD	p-value
Demographicsa
Age (years)	49.4 ± 13.9	49.6 ± 12.0	49.6 ± 12.3	0.952
Height (centimeters)	176.8 ± 7.7	168.9 ± 6.8	166.7 ± 8.8	0.000∗
Starting Weight (kg)	123.0 ± 24.8	118.2 ± 29.9	119.1 ± 28.9	0.591
Current Weight (kg)	111.4 ± 26.3	96.5 ± 25.9	99.6 ± 26.5	0.073
Starting BMIc (kg/mb)	39.6 ± 6.8	43.9 ± 11.0	43.0 ± 10.4	0.205
Current BMIc (kg/mb)	35.7 ± 7.5	35.9 ± 9.6	35.9 ± 9.1	0.945
Ethnicityb				0.412
Asian	1 (1.4 %)	0 (0 %)	1 (1.4 %)
African American	3 (4.3 %)	2 (2.9 %)	5 (7.2 %)
Hispanic/Latino	1 (1.4 %)	3 (4.3 %)	4 (5.8 %)
White/Caucasian	8 (11.6 %)	49 (71.0 %)	57 (82.6 %)
Education Levelb				0.727
Some high school	0 (0.0 %)	2 (2.9 %)	2 (2.9 %)
High school graduate or equivalent	3 (4.3 %)	10 (14.5 %)	13 (18.8 %)
Trade or vocational school degree	2 (2.9 %)	1 (1.4 %)	3 (4.3 %)
Some college	1 (1.4 %)	16 (23.2 %)	17 (24.6 %)
Associate's Degree	3 (4.3 %)	6 (8.7 %)	9 (13.0 %)
Bachelor's Degree	4 (5.8 %)	15 (21.7 %)	19 (27.5 %)
Graduate or professional degree	1 (1.4 %)	5 (7.2 %)	6 (8.7 %)

∗p ​< ​0.05 between genders.

^aDisplayed as Mean ± Standard Deviation.

^bDisplayed as count (percent) of the sample population.

^cBody Mass Index.

3.2. Healthy Eating Index

Results of the calculated 3-day food records revealed a total HEI score of 54 ± 12 (95 % CI 51.4, 57.3) with individual scores for the 13 components in Table 2. All HEI components, except added sugars, had statistically significant scores below the maximum possible scores (p < 0.001). Fig. 2 displays a radar plot with the HEI scores from this sample, the national average, and the perfect HEI score. One-way ANOVA analyses found no significant differences in HEI score for duration of GLP-1RA use (p = 0.266), ethnicity (p = 0.997), or education level (p = 0.340). A one-way ANOVA could not be performed for type of GLP-1RA due to only one person using Liraglutide. There was a non-significant, positive, weak relationship between current body weight and HEI score (r = 0.09, p = 0.461), and BMI (r = 0.173, p = 0.184).

Table 2.

Average Healthy Eating Index-2020 scores for a sample population of individuals using a glucagon-like-peptide-1 receptor agonist drug.

Component	Max Point	A n = 5	B n = 20	C n = 17	D n = 27	Total N = 69 Mean ± SD (95 % CI)	p
ADEQUACY COMPONENTS
Total Fruitsa	5	3.1	2.1	2.1	1.8	2.1 ​± ​1.9 (1.6, 2.5)∗	0.586
Whole Fruitsb	5	3.8	2.4	2.7	2.7	2.7 ​± ​2.2 (2.18, 3.2)∗	0.648
Total Vegetablesc	5	3.4	3.7	3.0	3.6	3.5 ​± ​1.5 (3.1, 3.8)∗	0.468
Greens and Beans	5	3.1	2.9	2.7	3.0	2.9 ​± ​2.1 (2.4, 3.4)∗	0.967
Whole Grains	10	2.9	2.1	2.5	2.5	2.4 ​± ​2.9 (1.7, 3.1)∗	0.926
Dairyd	10	5.8	6.7	5.0	6.1	6.0 ​± ​2.8 (5.3, 6.6)∗	0.352
Total Protein Foodsc	5	4.8	4.7	4.4	4.8	4.7 ​± ​0.7 (4.5, 4.9)∗	0.240
Seafood and Plant Proteinse	5	4.9	2.9	2.7	2.9	3.0 ​± ​2.2 (2.5, 3.5)∗	0.249
Fatty Acidsf	10	4.5	5.3	3.1	3.1	3.8 ​± ​3.2 (3.1, 4.6)∗	0.084
MODERATION COMPONENTS
Refined Grains	10	9.3	6.5	6.7	7.1	7.0 ​± ​3.3 (6.2, 7.8)∗	0.401
Sodium	10	2.7	2.6	3.3	2.3	2.7 ​± ​2.7 (2.0, 3.4)∗	0.723
Added Sugars	10	10	10	10	9.9	10 ± 0.2 (9.9, 10)	0.566
Saturated Fat	10	3.5	4.6	2.5	3.7	3.7 ​± ​8.4 (3.0, 4.4)∗	0.179
TOTAL SCORE	100	62 ± 9.7	56 ± 14	51 ± 12	54 ± 11	54 ​± ​12 (51.4, 57.3)∗#	0.266

Key: Duration of Medication Use Groups: A 1–3 months, B = 4–6 months, C = 7–12 months, and D = more than 1 year; Data displayed as Mean Scores; SD = Standard deviation; 95 % CI = 95 % Confidence interval lower bound, upper bound; p ​= ​between groups for duration of treatment; ∗ ​= ​p ​< ​0.001 and # ​= ​Bonferroni correction p ​< ​0.0038 compared to max score.

^aIncludes 100 % fruit juice.

^bIncludes all forms except juice.

^cIncludes beans, peas, and lentils.

^dIncludes all milk products, such as fluid milk, yogurt, and cheese, and fortified soy beverages.

^eIncludes seafood, nuts, seeds, soy products (other than beverages), and beans, peas, and lentils.

^fRatio of poly- and monounsaturated fatty acids to saturated fatty acids.

Fig. 2.

Radar plot displaying a perfect Healthy Eating Index Score, where all points touch the outer ring, compared to the National Average (ages 19–59) years old, and a sample of 69 individuals using a Glucagon-like-peptide 1 receptor agonist for weight loss.

3.3. Nutrient timing

Table 3 displays descriptive statistics for calories, macronutrients and fiber by meal and snack time. The largest calorie intake occurred at dinner time, as well as the highest intake of protein, fat, and carbohydrates compared to other mealtimes (see Table 3). Fig. 2 shows the percentage of calories from each macronutrient by meal and snack. Participants reported that 79 % of the food items were consumed at home with 93 %, 96 %, 97 %, and 84 % eating breakfast, lunch, dinner, and snacks, respectively. A Pearson's test found a strong, positive correlation between kcal and protein (r = 0.670, p < 0.001) and fiber (r = 0.662, p < 0.001). There was a very strong, positive relationship with kcal and fat (r = 0.890, p < 0.001) and carbohydrates (r = 0.873, p < 0.001). There were weak and moderate, positive correlations between current body weight and kcal (r = 0.298, p = 0.013), carbohydrates (r = 0.264, p = 0.029), protein (r = 0.347, p = 0.004), and fiber (r = 0.338, p = 0.004).

Table 3.

Average Calorie and Macronutrient Intake From 3-day food Records by Meals and Snacks for a Sample of Individuals using a Glucagon-like-peptide-1 receptor agonist drug.

	Breakfast (n = 64)	Lunch (n = 66)	Dinner (n = 67)	Snacks (n = 58)	Average Daily Total
Calories	392 ± 192 (341, 441)	538 ± 285 (468, 608)	649 ± 265 (584, 713)	391 ± 239 (329, 454)	1970
Protein (g)	16.3 ± 12 (13.2, 19.4)	24.7 ± 14.9 (21.0, 28.4)	33.5 ± 16.5 (29.5, 37.5)	10.2 ± 9.5 (7.7, 12.7)	84.7
Fat (g)	17.4 ± 13.4 (13.8, 20.8)	24.8 ± 15.1 (21.1, 28.5)	30.1 ± 14.5 (26.6, 33.6)	18.1 ± 13.7 (14.5, 21.7)	90.4
Carbohydrates (g)	43.3 ± 20.7 (37.9, 48.6)	54.5 ± 38.9 (43.4, 63.5)	61.1 ± 34.9 (52.6, 69.6)	48.0 ± 29.1 (40.4, 55.7)	206.9
Fiber (g)	3.5 ± 2.7 (2.8, 4.2)	4.0 ± 2.3 (3.5, 4.6)	5.6 ± 4.7 (4.4, 6.7)	4.1 ± 3.6 (3.2, 4.9)	17.2

Key: Reported as Mean ± Standard deviation (95 % Confidence interval lower bound, upper bound).

4. Discussion

The purpose of this secondary analysis was to determine the quality of dietary intake, independent of quantity, in individuals using a GLP-1RA. The analysis found dietary quality remains relatively poor regardless of calorie consumption. The sample had an average HEI score of 54 which aligns with previous research among adults [25]. Additionally, HEI scores did not differ significantly based on the duration of GLP-1RA use, there was a trend toward lower scores with GLP-1RA treatment. Compared to the maximum scores for each component of the HEI, the sample was significantly under the max scores for all components except added sugars. The sample had a perfect score for added sugars, meaning that less than 6.5 % of energy intake was from added sugars. This is aligned with the previous reports on grocery purchases showing that fewer sugary foods are purchased [28]. Within the adequacy components, whole grains and fatty acids had the biggest shortfalls with 76 % and 62 % under the max score, respectively. The total protein score was the closest to the max HEI score, however, the types of protein foods (seafood and plant proteins) were 40 % under the max score. For the moderation components, this sample consumed too much sodium and saturated fat compared to the guidelines with only 27 % and 37 % meeting the max score, respectively.

When matched for age with the national U.S. population [29] (NHANES 2017–2018,19–59 age group), the sample (average age = 50), scored a slightly lower total HEI of 54 compared to 57. Based on the adequacy components, the sample of individuals using a GLP-1RA had higher scores for vegetables (3.5 vs, 3.4), whole grains (2.4 vs. 2.3) and dairy (6.0 vs. 5.2) compared to U.S. population. However, this sample scored lower on total fruit (2.1 vs. 2.4), whole fruits (2.7 vs. 3.6), greens and beans (2.9 vs. 3.4), total protein (4.7 vs. 5.0), seafood and plant proteins (3.0 vs. 5.0), and fatty acids (3.8 vs. 4.4). In general, this sample consumes less fruit and protein from all sources compared to national averages. They are also consuming more saturated fat than unsaturated fat. Based on the moderation components, this sample had higher scores for refined grains (7.0 vs. 6.2) and added sugars (10 vs. 6.7) compared to the U.S. population. Whereas, they had lower scores for sodium (2.7 vs. 3.9) and saturated fat (3.7 vs. 5.2) compared to U.S. population. Based on this comparison, this sample consumes less refined grains and added sugars, but more sodium and saturated fat compared to the national U.S. population.

Compared to similar populations (e.g. medical weight reduction), individuals who underwent bariatric surgery had an average HEI-2015 of 50 and those who were eligible for the surgery scored an average of 48 from NHANES data 2015–2018 [30]. This sample of GLP-1RA scored higher than those who underwent bariatric surgery and those eligible for surgery by an average of 4 and 6 points, respectively. A meta-analysis from randomized weight reduction clinical trials found baseline HEI scores ranged from mid-30s to low 70s [31]. Interventions varied from prescribing particular foods to counseling with a dietitian with 4-to-7-point improvements most often found post-treatment. This implies interventions promoting behavioral changes can improve diet quality. A limitation of this study lacks the data on diet quality prior to beginning a GLP-1RA, therefore it's unknown if HEI scores have improved since beginning GLP-1RA treatments. Higher HEI scores are associated with lower risk of all cause death, cancer, and cardiovascular disease [32]. Future research and education are needed to improve HEI scores, especially with shifting dietary composition to include more fruits, vegetables, and protein.

This study also found the timing of nutrients across meals and snacks were suboptimal. Participants consumed the largest number of calories at dinner. Previous research has found that the largest meal reported at dinner is associated with higher BMI, whereas lunch as the largest meal was protective against obesity [33]. Further, having more than three meals per day was inversely associated with BMI. The sample within the study consumed the highest calories during dinner time. Based on this information, some recommendations for GLP-1 users should focus on disbursing meals throughout the day with the lowest calories at dinnertime. Research investigating meal timing strategies has found positive outcomes related to weight reduction and metabolic health. A large meta-analysis evaluating time-restricted eating, reduced meal frequency, or altering calorie distribution across the day found reduced weights compared to standard treatments [34]. Although the study populations were not using medical weight reduction drugs, lower meal frequency and earlier caloric intake were associated with greater weight reduction, suggesting this approach could complement GLP-1RA treatment. In the context of energy-reduced diets, front-loading caloric intake earlier in the day results in improved insulin resistance, fasting glucose, and LDL cholesterol [35]. Therefore, this pattern of distributing calorie intake to earlier in the day is advantageous for weight and metabolic health.

Protein dosing and distribution are key considerations during medical weight reduction therapy [36]. Based on the HEI protein component scores, protein intake both total protein and specifically seafood and plant proteins, were under the maximum score on average by 0.3 and 2 points. However, these maximum scores are calculated for general population needs and may not be a suitable measure for specialized conditions, such as medical weight reduction. Several publications emphasize the importance of high-protein diets during weight reduction [[37], [38], [39]]. To help preserve lean mass during hypocaloric diets, 1.2–2.0 g/kg of protein should be consumed daily based on adjusted body weight [36]. Evidence in clinical trials suggests distributing daily protein intake evenly across meals helps with preserving lean mass, which is a major concern during rapid weight reduction [40]. Evenly spreading protein intake throughout the day versus skewing protein intake towards the evening meal stimulates skeletal muscle protein synthesis to a greater degree [15]. In this sample, protein intake was skewed at dinner with 40 % of total protein intake consumed at dinner time with only 19 % at breakfast. Previous research shows better adherence to a weight reduction diet when eating 30 g of protein at each meal when trying to lose weight [41]. Guidance on protein intake for individuals using a GLP-1RA should emphasize adequate high-quality protein intake spread throughout the day with a goal of 30 g per meal and snack to preserve muscle mass. Consuming high amounts of protein at breakfast or more in the morning is associated with an increase in skeletal muscle index and lean body mass [42]. Therefore, guidelines for individuals using a GLP-1RA should be distributing protein intake throughout the day with a larger amount of protein consumed earlier in the day.

While more research is needed, results provide some initial insights into dietary patterns and nutrient timing in community-dwelling individuals using a GLP-1RA. Given the exploratory nature of the secondary analysis, large interpretation and conclusions should be drawn with caution. In general, nutritional guidance for GLP-1RA users should continue to focus on improving dietary quality and nutrient timing to help promote long-term health. Emphasis should be on increasing HEI adequacy components and reducing HEI moderation components. More specifically, more fruits, vegetables, whole grains, and healthy proteins should be included. Due to poor appetites, dietary supplements may support some of these adequacy goals, such as a fish oil supplement when protein sources from seafood and plant proteins remain low. Regarding nutrient timing, adequate protein intake should be evenly spread throughout the day to support lean mass. Protein supplements and meal replacement shakes can be a convenient source of protein to meet higher needs and support body composition [43]. In addition, reducing overall fat intake with an emphasis on unsaturated fats is particularly important. Lowering total fat intake and increasing fiber intake may also alleviate gastrointestinal side effects commonly caused by the usage of GLP-1RA medications [44]. The HEI does not measure water intake as a component of dietary quality, and future investigations should understand hydration status while using a GLP-1RA. Healthcare providers and dietitians can guide patients to use the adequacy and moderation components of the HEI as a simple tool to identify which foods to consume more of and which to limit with an aim to improve dietary quality. Future research is needed to determine if meaningful improvements in HEI scores can be a result of nutrition education interventions. More research is needed on optimal diet quality composition and specific nutrient and protein dosing and timing for the unique needs of the GLP-1RA population.

5. Limitations

The potential limitations of the current study include the following: 1) The convenience sample selected may not represent all individuals taking a GLP1-RA. However, this study provides initial insights on dietary pattern quality and nutrient timing for the population. 2) The data analyzed relied on self-reported data, assuming participants accurately reported dietary intakes and mealtimes. However, the ASA24 software provides strength in data collection for a 3-day food record as it provides visual examples of portion sizes, and probes for additional information to accurately gather nutrient intake and requires the participant to enter the mealtimes. 3) This study included participants using any type of GLP1-RA for various lengths spanning from 1 month to over one year. Different medication forms such as semaglutide versus tirzepatide could result in different nutrient intakes. 4) Further, additional factors that influence obesity management, such as the onset and duration of obesity, access to nutrition counseling, socioeconomic status, cultural influence, and food security should be addressed in future studies. It's currently unknown how dietary patterns change with long-term GLP-1RA treatments. The exploratory nature of the secondary analysis highlights the need for continued research on dietary quality, nutrient intake, and hydration status combined with GLP-1RA treatments.

6. Conclusion

This secondary analysis aimed to examine dietary quality and timing of nutrient intake with individuals using a GLP-1RA for obesity treatment. Within this sample, dietary quality based on validated HEI score was poor and slightly worse than the national average. However, in this sample, it is unknown whether the observed dietary patterns were pre-existing or potentially influenced by the GLP-1RA medication. Protein intake was skewed towards dinner rather than evenly distributed throughout the day and the largest calorie intake occurred at dinner. Future studies are needed to determine if other samples of individuals using a GLP-1RA medications find similar HEI score and intervention studies to explore ways to improve dietary quality and optimize nutrient intake and timing while also assessing the impact on health outcomes.

6.1. Key takeaway clinical messages

• •Individuals currently using a GLP-1RA for weight reduction demonstrated suboptimal dietary quality with low intake across adequacy components of the HEI including total and whole fruits, total vegetables and greens and beans, whole grains, seafood and plant proteins, dairy and fatty acids with excess intake of moderation components of the HEI including saturated fat, refined grains, sodium, and fatty acids among individuals currently using a GLP-1RA. However, the intake of added sugars met the recommended moderation limits.
• •Nutrient intake was skewed at dinner, which accounted for the largest intake of calories (average 649 kcal) and 40 % of total dietary protein intake, indicating suboptimal protein intake distributed throughout the day.
• •Fundamental nutritional recommendations for patients treated with GLP-1 RA include increasing fruits, vegetables, protein, whole grains, and fatty acid ratios, limiting sodium and saturated fats, and promoting adequate protein per meal and evenly distributed protein intake throughout the day, for a total of ∼100 g or more of protein daily.

CRediT author statement

Conceptualization and Methodology, B.J., M.M., L.G., R.K., R.J.; Data collection, B.J., Data analysis, B.J.; Validation, B.J., M.M., L.G.; Writing – original draft, B.J., M.M.; Writing – review & editing, B.J., M.M., L.G., R.K., R.J. All authors approved the final version.

Ethics review

Study procedures were approved by an Institutional Review Board, BRANY IRB-Manager, New York, USA (Approval date 16 August 2024), and all participants provided informed consent prior to voluntary participation.

Declaration of artificial intelligence (AI) and AI-assisted technologies utilized in the writing process

AI-technology was not used in the writing process.

Source of funding

This study was funded by GNC Holdings, LLC.

Declaration of competing interests

All authors are employed full-time by GNC Holdings, LLC.

Data availability

Data described in the manuscript code will be made available upon request.